Re: [PATCH v5 08/10] arm64: dts: qcom: Add initial support for MSM8917

[Stephan Gerhold <stephan.gerhold@xxxxxxxxxx>]

On Tue, Nov 12, 2024 at 04:49:38PM +0100, Barnabás Czémán wrote:
> From: Otto Pflüger <otto.pflueger@xxxxxxxxx>
> 
> Add initial support for MSM8917 SoC.
> 
> Signed-off-by: Otto Pflüger <otto.pflueger@xxxxxxxxx>
> [reword commit, rebase, fix schema errors]
> Signed-off-by: Barnabás Czémán <barnabas.czeman@xxxxxxxxxxxxxx>
> ---
>  arch/arm64/boot/dts/qcom/msm8917.dtsi | 1974 +++++++++++++++++++++++++++++++++
>  1 file changed, 1974 insertions(+)
> 
> diff --git a/arch/arm64/boot/dts/qcom/msm8917.dtsi b/arch/arm64/boot/dts/qcom/msm8917.dtsi
> new file mode 100644
> index 0000000000000000000000000000000000000000..cf0a0eec1141e11faca0ee9705d6348ab32a0f50
> --- /dev/null
> +++ b/arch/arm64/boot/dts/qcom/msm8917.dtsi
> @@ -0,0 +1,1974 @@
> [...]
> +		domain-idle-states {
> +			cluster_sleep_0: cluster-sleep-0 {
> +				compatible = "domain-idle-state";
> +				arm,psci-suspend-param = <0x41000023>;
> +				entry-latency-us = <700>;
> +				exit-latency-us = <650>;
> +				min-residency-us = <1972>;
> +			};
> +
> +			cluster_sleep_1: cluster-sleep-1 {
> +				compatible = "domain-idle-state";
> +				arm,psci-suspend-param = <0x41000043>;
> +				entry-latency-us = <240>;
> +				exit-latency-us = <280>;
> +				min-residency-us = <806>;
> +			};

I think my comment here is still open:

This is strange, the deeper sleep state has lower timings than the
previous one?

> +
> +			cluster_sleep_2: cluster-sleep-2 {
> +				compatible = "domain-idle-state";
> +				arm,psci-suspend-param = <0x41000053>;
> +				entry-latency-us = <700>;
> +				exit-latency-us = <1000>;
> +				min-residency-us = <6500>;
> +			};
> +		};
> +
> [...]
> +		restart@4ab000 {
> +			compatible = "qcom,pshold";
> +			reg = <0x004ab000 0x4>;
> +		};

This one too:

You have PSCI for shutting down, do you actually need this?

> +
> +		tlmm: pinctrl@1000000 {
> +			compatible = "qcom,msm8917-pinctrl";
> +			reg = <0x01000000 0x300000>;
> +			interrupts = <GIC_SPI 208 IRQ_TYPE_LEVEL_HIGH>;
> +			gpio-controller;
> +			gpio-ranges = <&tlmm 0 0 134>;
> +			#gpio-cells = <2>;
> +			interrupt-controller;
> +			#interrupt-cells = <2>;
> +
> [...]
> +			sdc1_clk_on: sdc1-clk-on-state {
> +				pins = "sdc1_clk";
> +				bias-disable;
> +				drive-strength = <16>;
> +			};
> +
> +			sdc1_clk_off: sdc1-clk-off-state {
> +				pins = "sdc1_clk";
> +				bias-disable;
> +				drive-strength = <2>;
> +			};
> +
> +			sdc1_cmd_on: sdc1-cmd-on-state {
> +				pins = "sdc1_cmd";
> +				bias-disable;
> +				drive-strength = <10>;
> +			};
> +
> +			sdc1_cmd_off: sdc1-cmd-off-state {
> +				pins = "sdc1_cmd";
> +				bias-disable;
> +				drive-strength = <2>;
> +			};
> +
> +			sdc1_data_on: sdc1-data-on-state {
> +				pins = "sdc1_data";
> +				bias-pull-up;
> +				drive-strength = <10>;
> +			};
> +
> +			sdc1_data_off: sdc1-data-off-state {
> +				pins = "sdc1_data";
> +				bias-pull-up;
> +				drive-strength = <2>;
> +			};
> +
> +			sdc1_rclk_on: sdc1-rclk-on-state {
> +				pins = "sdc1_rclk";
> +				bias-pull-down;
> +			};
> +
> +			sdc1_rclk_off: sdc1-rclk-off-state {
> +				pins = "sdc1_rclk";
> +				bias-pull-down;
> +			};
> +
> +			sdc2_clk_on: sdc2-clk-on-state {
> +				pins = "sdc2_clk";
> +				drive-strength = <16>;
> +				bias-disable;
> +			};
> +
> +			sdc2_clk_off: sdc2-clk-off-state {
> +				pins = "sdc2_clk";
> +				bias-disable;
> +				drive-strength = <2>;
> +			};
> +
> +			sdc2_cmd_on: sdc2-cmd-on-state {
> +				pins = "sdc2_cmd";
> +				bias-pull-up;
> +				drive-strength = <10>;
> +			};
> +
> +			sdc2_cmd_off: sdc2-cmd-off-state {
> +				pins = "sdc2_cmd";
> +				bias-pull-up;
> +				drive-strength = <2>;
> +			};

These are not referenced anywhere? Not here in the sdhc_X nodes, and
also not in your msm8917-xiaomi-riva.dts. Would also recommend
consolidating these to a single node like in msm8916.dtsi, see commit
c943e4c58b2f ("arm64: dts: qcom: msm8916/39: Consolidate SDC pinctrl").

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=c943e4c58b2ffb0dcd497f8b12f284f5e8fc477e

> +
> +			sdc2_cd_on: cd-on-state {
> +				pins = "gpio67";
> +				function = "gpio";
> +				drive-strength = <2>;
> +				bias-pull-up;
> +			};
> +
> +			sdc2_cd_off: cd-off-state {
> +				pins = "gpio67";
> +				function = "gpio";
> +				drive-strength = <2>;
> +				bias-disable;
> +			};

It does not make sense to have different on/off states for the card
detect (CD) pin of the SD card. It needs to work even when the SD card
is suspended so we can detect insertions/removals. Also should be placed
in the board-specific DT part.

See commit dfbda20dabaa ("arm64: dts: qcom: msm8916/39: Fix SD card
detect pinctrl").

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=dfbda20dabaa1f284abd550035db5887384c8e4c


> +
> +			sdc2_data_on: sdc2-data-on-state {
> +				pins = "sdc2_data";
> +				bias-pull-up;
> +				drive-strength = <10>;
> +			};
> +
> +			sdc2_data_off: sdc2-data-off-state {
> +				pins = "sdc2_data";
> +				bias-pull-up;
> +				drive-strength = <2>;
> +			};
> +
> [...]
> +		blsp1_i2c4: i2c@78b8000 {
> +			compatible = "qcom,i2c-qup-v2.2.1";
> +			reg = <0x078b8000 0x500>;
> +			interrupts = <GIC_SPI 98 IRQ_TYPE_LEVEL_HIGH>;
> +			clocks = <&gcc GCC_BLSP1_QUP4_I2C_APPS_CLK>,
> +				 <&gcc GCC_BLSP1_AHB_CLK>;
> +			clock-names = "core", "iface";
> +			dmas = <&blsp1_dma 10>, <&blsp1_dma 11>;
> +			dma-names = "tx", "rx";
> +			pinctrl-0 = <&blsp1_i2c4_default>;
> +			pinctrl-1 = <&blsp1_i2c4_sleep>;
> +			pinctrl-names = "default", "sleep";
> +			#address-cells = <1>;
> +			#size-cells = <0>;
> +			status = "disabled";
> +		};
> +
> +		blsp2_i2c5: i2c@7af5000 {

This is actually blsp2_i2c1 if you look at the clock name below:

> +			compatible = "qcom,i2c-qup-v2.2.1";
> +			reg = <0x07af5000 0x600>;
> +			interrupts = <GIC_SPI 299 IRQ_TYPE_LEVEL_HIGH>;
> +			clocks = <&gcc GCC_BLSP2_QUP1_I2C_APPS_CLK>,

here ^

But I realize now that the pinctrl functions are consecutively numbered
without the BLSP number. Sorry for the confusion.

Basically:
  - blsp1_i2c2 == blsp_i2c2
  - blsp2_i2c1 == blsp_i2c5

Looking at some other examples upstream I guess you can choose between
one of the following options:

 1. msm8974/msm8976/msm8996/msm8998: Use &blspX_i2cY labels for the i2c@
    node and pinctrl and only have the slightly confusing pinctrl
    function. E.g. this in msm8976.dtsi:

			/* 4 (not 6!) interfaces per QUP, BLSP2 indexes are numbered (n)+4 */
			blsp2_i2c2_default: blsp2-i2c2-default-state {
				pins = "gpio22", "gpio23";
				function = "blsp_i2c6";
				drive-strength = <2>;
				bias-disable;
			};

    Note how blsp2_i2c2 == blsp_i2c6.

 2. msm8994: Use &blspX_i2cY labels for the i2c@ node, but keep pinctrl
    named &i2cN_default. E.g. this in msm8994.dtsi:

		blsp2_i2c1: i2c@f9963000 {
			/* ... */
			pinctrl-names = "default", "sleep";
			pinctrl-0 = <&i2c7_default>;
			pinctrl-1 = <&i2c7_sleep>;
			/* ... */
		};

    Note how blsp2_i2c1 == i2c7_default here.

 3. msm8953: Use &i2c_N labels everywhere like on downstream. E.g. this
    in msm8953.dtsi. This is pretty much what you had originally:

		i2c_5: i2c@7af5000 {
			/* ... */
			pinctrl-names = "default", "sleep";
			pinctrl-0 = <&i2c_5_default>;
			pinctrl-1 = <&i2c_5_sleep>;
			/* ... */
		};

All of these are fine for me. Feel free to pick the one you prefer. But
let's not introduce a new confusing variant of this. :-)

Thanks,
Stephan